Skip to content

muxover/js-view

v0.1.0MIT

Read JavaScript-rendered pages with a real browser: render, screenshot, links, and API-call capture as MCP tools, plus a skill that tells the agent when to use them.

GitHubOpen Website

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

0.1.0 - 2026-10-03

Added

  • js-view CLI: render a page to markdown, text, HTML, or JSON, list its links, capture its XHR/fetch calls, or save a screenshot.
  • MCP server (js-view mcp) with render, screenshot, links, and network tools; the browser stays warm between calls.
  • Plugins for Claude Code, Codex, and Cursor, a one-click Cursor MCP install, and a skill that installs in any agent with npx skills add muxover/js-view.
  • Render outcomes: bot checks, login walls, HTTP errors, and empty pages come back labelled with hints instead of passed off as content.
  • selector to keep only matching elements and max_chars to cut long content at a paragraph.
  • Page actions: scroll, click, type, wait, and navigate.
  • js-view login to sign in once in a visible browser and reuse the session.
  • Uses Playwright's Chromium, or an installed Chrome or Edge; js-view install downloads Chromium when neither is present.
  • Screenshots of the viewport, the full page, or one element, with optional OCR.
  • Self-hosted HTTP service with a browser pool, API-key auth, rate limiting, an SSRF guard for private hosts, and Redis-backed workers.